Search

In this paper, we propose a novel and enhanced approach for crowd counting within the domain of manatee monitoring, aiming to significantly improve efficiency and accuracy. The proposed model achieves state-of-the-art results in the challenging task of manatee counting, simplifying the work of scientists and experts in the field. Our model not only facilitates the identification and enumeration of manatees in images and videos but also excels in scenarios that pose considerable challenges for human observers. To enhance accurate counting of the manatee aggregation, we introduce a framework with three key innovations to tackle the challenge: a new approach to generate density maps during the training process, an augmented technique to balance the dataset, and a cross-domain solution to enhance overall performance. The proposed two-dimensional Gaussian kernel offers a refined method for creating density maps, providing a more robust foundation for the training phase. Additionally, we built a balanced and augmented dataset, ensuring that the model is exposed to diverse and representative instances, thus improving its generalization capabilities. Furthermore, we incorporate a cross-domain phase pretraining the model utilizing an image dataset of wild animals to initialize the weights and further improve performance. Experiments and comparisons, with respect to previously established CSRNET model presented in Wang et al. (2023), demonstrate noteworthy improvements. Remarkably, our model achieves a Mean Absolute Error (MAE) of nearly half compared to the rival approach, showcasing the substantial advancements achieved through our refined methodology. This progress boosts the reliability of manatee counting in conservation efforts and ecological research.

Evidence showed that N6-methyladenosine (m6A) modification plays a pivotal role in influencing RNA fate and is strongly associated with cell growth and developmental processes in many species. However, no information regarding m6A modification in Eimeria tenella is currently available. In the present study, we surveyed the transcriptome-wide prevalence of m6A in sporulated oocysts and unsporulated oocysts of E. tenella. Methylated RNA immunoprecipitation sequencing (MeRIP-seq) analysis showed that m6A modification was most abundant in the coding sequences, followed by stop codon. There were 3,903 hypermethylated and 3,178 hypomethylated mRNAs in sporulated oocysts compared with unsporulated oocysts. Further joint analysis suggested that m6A modification of the majority of genes was positively correlated with mRNA expression. The mRNA relative expression and m6A level of the selected genes were confirmed by quantitative reverse transcription PCR (RT-qPCR) and MeRIP-qPCR. GO and KEGG analysis indicated that differentially m6A methylated genes (DMMGs) with significant differences in mRNA expression were closely related to processes such as regulation of gene expression, epigenetic, microtubule, autophagy-other and TOR signaling. Moreover, a total of 96 DMMGs without significant differences in mRNA expression showed significant differences at protein level. GO and pathway enrichment analysis of the 96 genes showed that RNA methylation may be involved in cell biosynthesis and metabolism of E. tenella. We firstly present a map of RNA m6A modification in E. tenella, which provides significant insights into developmental biology of E. tenella.

Abstract Manatees are aquatic mammals with voracious appetites. They rely on sea grass as the main food source, and often spend up to eight hours a day grazing. They move slow and frequently stay in groups (i.e. aggregations) in shallow water to search for food, making them vulnerable to environment change and other risks. Accurate counting manatee aggregations within a region is not only biologically meaningful in observing their habit, but also crucial for designing safety rules for boaters, divers, etc., as well as scheduling nursing, intervention, and other plans. In this paper, we propose a deep learning based crowd counting approach to automatically count number of manatees within a region, by using low quality images as input. Because manatees have unique shape and they often stay in shallow water in groups, water surface reflection, occlusion, camouflage etc. making it difficult to accurately count manatee numbers. To address the challenges, we propose to use Anisotropic Gaussian Kernel (AGK), with tunable rotation and variances, to ensure that density functions can maximally capture shapes of individual manatees in different aggregations. After that, we apply AGK kernel to different types of deep neural networks primarily designed for crowd counting, including VGG, SANet, Congested Scene Recognition network (CSRNet), MARUNet etc. to learn manatee densities and calculate number of manatees in the scene. By using generic low quality images extracted from surveillance videos, our experiment results and comparison show that AGK kernel based manatee counting achieves minimum Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). The proposed method works particularly well for counting manatee aggregations in environments with complex background.

Abstract The success of the business directly contributes towards the growth of the nation. Hence it is important to evaluate and predict whether the business will be successful or not. In this study, we use the company’s dataset which contains information from startups to Fortune 1000 companies to create a machine learning model for predicting business success. The main challenge of business success prediction is twofold: (1) Identifying variables for defining business success; (2) Feature selection and feature engineering based on Investor-Business-Market interrelation to provide a successful outcome of the predictive modeling. Many studies have been carried out using only the available features to predict business success, however, there is still a challenge to identify the most important features in different business angles and their interrelation with business success. Motivated by the above challenge, we propose a new approach by defining a new business target based on the definition of business success used in this study and develop additional features by carrying out statistical analysis on the training data which highlights the importance of investments, business, and market features in forecasting business success instead of using only the available features for modeling. Ensemble machine learning methods as well as existing supervised learning methods were applied to predict business success. The results demonstrated a significant improvement in the overall accuracy and AUC score using ensemble methods. By adding new features related to the Investor-Business-Market entity demonstrated good performance in predicting business success and proved how important it is to identify significant relationships between these features to cover different business angles when predicting business success. Graphical Abstract

Spatio-temporal forecasting is of great importance in a wide range of dynamic systems applications, such as earth science, transport planning, etc. These applications rely on accurate predictions of spatio-temporal structured data reflecting real-world phenomena. A stunning characteristic is that the dynamical system is not only driven by some physics laws but also impacted by the localized factor in spatial and temporal regions. One of the major challenges is to infer the underlying causes, which generate the perceived data stream and propagate the involved causal dynamics through the distributed observing units. Another challenge is that the success of machine learning-based predictive models requires massive annotated data for model training. However, the acquisition of high-quality annotated data is objectively manual and tedious as it needs a considerable amount of human intervention, making it infeasible in fields that require high levels of expertise. To tackle these challenges, we advocate a spatio-temporal physics-coupled neural networks (ST-PCNN) model to learn the underlying physics of the dynamical system and further couple the learned physics to assist the learning of the recurring dynamics. To deal with data-acquisition constraints, an active learning mechanism with Kriging for actively acquiring the most informative data is proposed for ST-PCNN training in a partially observable environment. Our experiments on both synthetic and real-world datasets exhibit that the proposed ST-PCNN with active learning converges to near-optimal accuracy with substantially fewer instances.

Federated Learning (FL), as a new computing framework, has received significant attentions recently due to its advantageous in preserving data privacy in training models with superb performance. During FL learning, distributed sites first learn respective parameters. A central site will consolidate learned parameters, using average or other approaches, and disseminate new weights across all sites to carryout next round of learning. The distributed parameter learning and consolidation repeat in an iterative fashion until the algorithm converges or terminates. Many FL methods exist to aggregate weights from distributed sites, but most approaches use a static node alignment approach, where nodes of distributed networks are statically assigned, in advance, to match nodes and aggregate their weights. In reality, neural networks, especially dense networks, have nontransparent roles with respect to individual nodes. Combined with random nature of the networks, static node matching often does not result in best matching between nodes across sites. In this paper, we propose, FedDNA, a dynamic node alignment federated learning algorithm. Our theme is to find best matching nodes between different sites, and then aggregate weights of matching nodes for federated learning. For each node in a neural network, we represent its weight values as a vector, and use a distance function to find most similar nodes, i.e., nodes with the smallest distance from other sides. Because finding best matching across all sites are computationally expensive, we further design a minimum spanning tree based approach to ensure that a node from each site will have matched peers from other sites, such that the total pairwise distances across all sites are minimized. Experiments and comparisons demonstrate that FedDNA outperforms commonly used baseline, such as FedAvg, for federated learning.

Toxoplasma gondii (T. gondii) is responsible for severe human and livestock diseases, huge economic losses, and adversely affects the health of the public and the development of animal husbandry. Glycosylation is a common posttranslational modification of proteins in eukaryotes, and N-glycosylation is closely related to the biological functions of proteins. However, glycosylation alterations in the feline small intestine following T. gondii infection have not been reported. In this study, the experimental group was intragastrically challenged with 600 brain cysts of the Prugniuad (Pru) strain that were collected from infected mice. The cats’ intestinal epithelial tissues were harvested at 10 days post-infection and then sent for protein glycosylation analysis. High-performance liquid chromatography coupled to tandem mass spectrometry was used to analyze the glycosylation alterations in the small intestine of cats infected with T. gondii. The results of the present study showed that 56 glycosylated peptides were upregulated and 37 glycosylated peptides were downregulated in the feline small intestine infected by T. gondii. Additionally, we also identified eight N-glycosylated proteins of T. gondii including eight N-glycopeptides and eight N-glycosylation sites. The protein A0A086JND6_TOXGO (eEF2) and its corresponding peptide sequence were identified in T. gondii infection. Some special GO terms (i.e., cellular process and metabolic process, cell and cell part, and catalytic activity) were significantly enriched, and the Clusters of Orthologous Groups of proteins (COG) function prediction results showed that posttranslational modification, protein turnover, and chaperones (11%) had the highest enrichment for T. gondii. Interestingly, eEF2, a protein of T. gondii, is also involved in the significantly enriched T. gondii MAPK pathway. The host proteins ICAM-1 and PPT1 and the endoplasmic reticulum stress pathway may play an important role in the glycosylation of Toxoplasma-infected hosts. This is the first report showing that T. gondii oocysts can undergo N-glycosylation in the definitive host and that eEF2 is involved, which may provide a new target for T. gondii detection to prevent the spread of T. gondii oocysts in the future.

Eimeria tenella encodes a genome of approximately 8000 genes. To date, however, very few data are available regarding E. tenella rhoptry kinase family proteins. In the present study, the gene fragment encoding the mature peptide of the rhoptry kinase family protein 17 of E. tenella (EtROP17) was amplified by PCR and expressed in E. coli. Then, we generated polyclonal antibodies that recognize EtROP17 and investigated the expression of EtROP17 in the merozoite stage of E. tenella by immunofluorescent staining and Western blot analysis. Meanwhile, the protective efficacy of rEtROP17 against E. tenella was evaluated in chickens. Sequencing analysis showed that a single base difference at sequence position 1901 was observed between the SD-01 strain and the Houghton strain. EtROP17 was expressed in the merozoite stage of E. tenella. The results of the animal challenge experiments demonstrated that vaccination with rEtROP17 significantly reduced cecal lesions and oocyst outputs compared with the challenged control group. Our findings indicate that EtROP17 could serve as a potential candidate for developing a new vaccine against E. tenella.